In-vivo RNA and DNA polymerases utilize ribonucleotides and deoxyribonucleotides respectively, for the polymerisation of nucleic acids, and discriminate between these nucleotides with high fidelity. Extensive efforts have been made to permit the incorporation of modified nucleotides by polymerases, including base modifications, sugar modifications, and backbone modifications. The use of such modified nucleotides as substrates for both RNA-and DNA-polymerases is desirable for a variety of reasons. Among others these include the incorporation of fluorescent labels for product detection (Raap), ribose-modified nucleotides for the generation of polynucleotides less susceptible to nuclease action (Sioud) or the use of terminating dideoxyribose nucleotides for DNA sequencing (Sanger).
Attempts to use modified nucleotides are often hampered by the substrate specificity of the polymerase in question. Modification with additional chemical moieties on the base have on the whole met with good success providing that the additional groups are attached to non Watson-Crick pairing residues and project out of the major groove. On the other hand modifications of the sugar rings have proven to be much less well tolerated presumably reflecting the presence of exquisitely precise interactions between the sugar and the enzyme during nucleotide binding and catalysis. Nevertheless there are several notable successes of the employment of mutagenesis to engineer polymerases capable of improved capacity to use sugar-modified nucleotides. In the case of DNA sequencing, polymerases have been improved in their utilisation of 2′3′-dideoxynucleotide terminators by engineering an amino acid substitution in the active site rendering the polymerase more similar to T7 DNA polymerase, demonstrated to tolerate such nucleotides well (Tabor).
Furthermore investigation into the biochemical and structural source of differing substrate specificity between bacterial and phage DNA and RNA polymerases has lead to the identification of residues which if substituted enable RNA polymerases to use deoxyribonucleotides, and others that allow DNA polymerases to use ribonucleotides. In particular a peptide loop referred to as the ‘steric gate’ appears to prevent DNA polymerases from accepting groups bulkier than the lone hydrogen atom present at the 2′ position in deoxyribonucleotides. This loop is essentially missing in phage-encoded RNA polymerases. On this basis there is reason to believe that RNA polymerases may be somewhat more tolerant to the presence of other groups attached to the 2′ position of the sugar ring providing the groups are relatively small. Consistent with this Padilla and Sousa have shown that T7 RNA polymerase is capable of utilising nucleotides modified at the 2′ position with O-methyl groups, or with azido (N3) groups (Padilla), provided that one or two additional enabling amino acid substitutions are also introduced.